A STATISTICAL STUDY OF MATE CHOICE: SEXUAL SELECTION IN A PLETHODONTID SALAMANDER (DESMOGNATHUS OCHROPHAEUS)

Our experiment revealed the existence of significant variation in mating success in a salamander species in which males do not provide courtship feeding, nest sites, or parental care. Differences in mating success were based on natural variation among adult males and females, rather than on traits of an artificially selected set of potential mates. Courtship encounters deliberately involved only one male and one female, thus eliminating the potentially confounding effects of male-male competition and variations in mate encounter rate. Even after eliminating these effects and random error, some females were more likely than others to elicit spermatophore deposition by a male, and some males were more likely than others to inseminate a female. Such variation among individuals represents an opportunity for sexual selection to act on phenotypic characters that affect mating success. We advocate the use of a factorial experimental design to analyze sexual selection. This approach permits the statistical evaluation of separate male and female effects, interaction between these effects, and random effects. Designs which combine the evaluations of mating success and courtship behaviors could estimate the force of sexual selection on behavior.